Meta has postponed the launch of its next-generation AI model, code-named Avocado, from March to at least May 2026 after internal testing revealed it underperforms leading systems from OpenAI, Anthropic, and Google in reasoning, coding, and writing tasks. The model was intended as a major reset after the lukewarm reception of Llama 4, but benchmarks placed it between Google's Gemini 2.5 and Gemini 3.0 — short of the frontier performance Meta had targeted.
The delay puts growing pressure on Meta's $115–135 billion AI capital spending plan for 2026, the largest in the company's history. Unlike Amazon, Microsoft, and Google, Meta lacks a cloud business to monetize its AI infrastructure directly, making product-integrated AI across WhatsApp, Instagram, and Facebook critical to justifying the investment. Meta's leadership has reportedly discussed temporarily licensing Google's Gemini technology as a stopgap, though no decision has been made.
